Experienced Forum User, Published Author, Player
(20)
Joined: 10/14/2005
Posts: 317
Indeed, but the number of item1s seems to have a pretty direct effect on lag, while their placement does not. With the laggy spot in Airman stage I found it was possible to scroll with two item1s, but it became consistently easier with three item1s.
A crash bomb is lag-equivalent to about how many item1s? Also, do some of the weapons which can be used three at a time (metal blades, air tornadoes, and quick boomerangs) slow the game down any significant amount?
Experienced Forum User, Published Author, Player
(20)
Joined: 10/14/2005
Posts: 317
The item1s don't have to be in the same place: shuttertech-crashman-separateitem1.fcm, and throwing them together doesn't seem to actually increase the probability of success. However, I have yet to perform the trick with less than three item1s.
Experienced Forum User, Published Author, Player
(20)
Joined: 10/14/2005
Posts: 317
I'm not getting very far trying to reproduce this, but here's one useless characteric I've found playing off Shinryuu's movie:
If the B button is held at the point of screen edge contact, this disables the technique (even though it doesn't seem to affect the lag), while it is unaffected by any random B button presses or releases just before that point of contact. Other extra button presses like up, down, select, start, A in this interval do not change the result, as long as the maneuver remains the same.
Experienced Forum User, Published Author, Player
(20)
Joined: 10/14/2005
Posts: 317
Well, the ceiling doesn't have any overlapping blocks so there's nothing to separate out. In case it might have something to do with the ceiling block type (e.g. the jagged rather than flat edge), here are two more test cases:
smb3_pyramid_nojagged.ipssmb3_pyramid_coin.ips
Experienced Forum User, Published Author, Player
(20)
Joined: 10/14/2005
Posts: 317
I took a look at that spot in the level editor and one strange thing about it is that there are actually two blocks superimposed to make up that wall. Perhaps this forces Mario out with twice as much force, making it a lot harder to duck-jump through it?
I've made an IPS patch that separates out the extra wall so that maybe someone here could test this theory: smb3_pyramidwall.ips
Experienced Forum User, Published Author, Player
(20)
Joined: 10/14/2005
Posts: 317
Press the comma , key to turn on input display. That's how to quickly confirm that turbo is working properly.
Double Dragon and probably some other games require the A or B buttons to be released for 2 frames before being pressed again in order to count for an additional punch/kick.
Implementing an option to allow the turbo function to use a configurable number of frames would solve this, but I don't think there's much demand for such a feature. What's the point of using turbo in a game like Double Dragon?
Experienced Forum User, Published Author, Player
(20)
Joined: 10/14/2005
Posts: 317
OmnipotentEntity wrote:
A very easy way to test this is make a movie with DK beating the first level in time attack. Then hex it to a movie where you selected Bowser instead. If the it doesn't desync and the time is the same then they are the same character.
Unfortunately that isn't very easy nor would it prove conclusive. Since this game suffers from sporadic lag bubbles, any minor change could quickly cause a desync regardless.
Frame-counting is an easier way to test it. For the initial starts on various races, I've confirmed what Saturn said:
Mario-Luigi: same
Princess-Yoshi: same
Bowser-DK: same
Koopa-Toad: same
There aren't any races that have a long enough straightaway in the beginning to simply reach top speed for Bowser/DK, but these tests already shows that the Wikipedia article is wrong on the various characters' accelerations.
Experienced Forum User, Published Author, Player
(20)
Joined: 10/14/2005
Posts: 317
Ok. Here you go: Legend of Zelda, The - A Link to The Past (U) [!].smv. I didn't bother optimizing all the movements.
Considering players can do this on the console consistently, I don't think there's much point in doing it tool-assisted. Perhaps a more interesting challenge would be to make a video that shows how to obtain the room using 4 to 5 of the different possible routes.
Experienced Forum User, Published Author, Player
(20)
Joined: 10/14/2005
Posts: 317
I was fighting armored Vile for no good reason, and I noticed that when you hit him with ice he ends up facing the opposite direction while being sucked in towards X. For kicks, this can even make him run backwards or shoot fireballs backwards. Maybe this could be used to draw him in to get damaged a couple frames earlier, and might also help with placing Vile in the right spot at the end of the battle.
edit: Looking at Dehacked's movie, I watched further on to the boss fights and wondered about some possible improvements, particularly involving charged weapons. FractalFusion's 100% test run does try out various boss fight strategies involving charging, but so far they've all looked slower to me. Here are some ways that I think fights could actually be made faster. I'm sure some of them might already have been brought up.
- The shotgun ice appears to be slightly faster than the regular shot, although I can't tell for sure. If it is, you could save maybe 1 frame by using it to start off a fight in a battle that can afford the 1 unit of damage slack. In Dehacked's movie, only the second battle with Chill Penguin seems to fit this category.
- Boomer Kuwanger (second encounter): First fire a regular missile while using the bug to continue charging. Then jump around and fire a charged four-missile blast and time it so that Boomer Kuwanger gets hit by one, dashes away, then gets hit by one of the other charged missiles. Then finish the battle and it should be faster despite a little time lost by starting with a missile instead of a fast regular shot.
- Armored Armadillo (second encounter in any% run and first encounter in 100% run) strategy: Was this already mentioned? Start off with a charged spark, then use a regular spark, then get another charged spark ready in time by taking advantage of the waiting time during his electrocution animation.
- Sting Chameleon (second encounter in any% run and first encounter in 100% run). I think FractalFusion's test run demonstrated this somewhat, although not perfectly. It's theoretically possible to get a boomerang to orbit X indefinitely, so set that up early in order to buy some charging in order to fire a second charged boomerang blast without losing time.
Experienced Forum User, Published Author, Player
(20)
Joined: 10/14/2005
Posts: 317
Phil wrote:
For timeattacks a keyboard is better.
This depends on where your comfort zone is and how many hotkeys you actually need. I found it useful to assign gamepad buttons for pause, frame-advance, load, and save. I think Randil has mentioned that he also does this. A device with a lot of buttons like the GameCube controller would work well in such a setup, while on the other hand a NES controller wouldn't make much sense due to its lack of buttons.
Lately I've used an SNES controller for all my gaming and TAS stuff.
SNES: Here is what I have, or what Shinryuu and FODA wish they had.
PlayStation: Here is just one example of the adapters used by Bablo, Zoizite, and FODA. I bought this particular one back when it was known as the Super Dual Box.
Experienced Forum User, Published Author, Player
(20)
Joined: 10/14/2005
Posts: 317
AngerFist wrote:
If you would check his last 30-40 posts, you would agree with me that many of them are utterly pointless posts.
So you convinced me to try checking his last 30-40 posts. I only got up to the last 15, but here is the count so far:
- 10 with direct messages (getting his points across often in just one line, which is a plus)
- 2 funny with hidden messages (including one in this thread)
- 3 funny but pointless
This could even be further offset by my inability to see more of the hidden messages. I think you'd get a similar ratio if you counted them up to the last 40.
Omega wrote:
stuff
Yeah I'm sure you'd think that with a name like Omega! Of course I'm inclined to agree with a name like mine. There are already a few hanzous in some well-known video games and movies, and plenty living in Japan today. I'm just thankful this was overlooked when my movie was published a couple weeks back.
Bisqwit wrote:
But I'm not a fan of anonymous submissions either, because if you are afraid of staining your name with something that you've submitted and thus don't want to include a name that identifies you, perhaps you shouldn't submit it in the first place.
Consider this. Why does Batman use just his nick and symbol instead of at least some more descriptive abbreviation like BruceW or BWayne? Yes, my faulty analogy here is suggesting to compare TAS-makers with superheroes.
Regarding the submission, I watched the movie and voted meh, which would be the most generous vote so far. To me, letting enemies get away is what stands out most to make it look sloppy.
Experienced Forum User, Published Author, Player
(20)
Joined: 10/14/2005
Posts: 317
FODA had suggested grabbing the wand directly from a wall jump in order to catch it as high as possible. Did you try that? I see you do a little wall jump but then end up jumping off the ground to catch the wand, so is the problem that there's not enough momentum when bouncing off the wall?
Experienced Forum User, Published Author, Player
(20)
Joined: 10/14/2005
Posts: 317
My impression from playing on the console is that you might lose even more time by fighting Boomer Kuwanger with the X-buster/shotgun ice, but good luck to you testing that.
Have you guys measured the exact frame counts for the first and second Storm Eagle battles on Dehacked's run? Unless either fight is significantly improvable, this difference ought to give a good measurement as to how much can be gained from these order changes.
Experienced Forum User, Published Author, Player
(20)
Joined: 10/14/2005
Posts: 317
So how much time is lost by fighting the serpent with the next best weapon? I have a feeling it might even be more than five seconds, in which case this route might not be worth exploring.
Experienced Forum User, Published Author, Player
(20)
Joined: 10/14/2005
Posts: 317
Well, both the old and new versions of FCEU seem to save the configuration only when the program is exited normally. Users who have the habit of leaving one instance of a program open for days on end are more likely to finish those sessions with FCEU crashing some way or another, in which case any hotkey settings made long ago can still be lost.
I'd prefer if settings were always saved immediately after making changes to any of the things in the Config menu.
Experienced Forum User, Published Author, Player
(20)
Joined: 10/14/2005
Posts: 317
Sleepz wrote:
I’ve tried doing the potion trick on all the different versions of the game with no success, so I guess we can scratch that idea. Glad I didn’t go through with the run. I did find another way to pointlessly freeze the game though, this time with left+right.
Experienced Forum User, Published Author, Player
(20)
Joined: 10/14/2005
Posts: 317
Matty, that movie is in Nesticle and it's been around for quite a while. You can download it here. I don't think it's really cool by our TAS standards though.
Food's right about swordless vs sword-abstinent. I made the same point in some other thread, and I'm still waiting for you guys to tell me which Nintendo Power issue that was.
Yes you can use up+A in Ganon's room, and I think a few minutes added onto a half-hour run would be worth making it much closer to swordless.
Last week I tried to think up a better solution by maybe bending one of the site rules. I figured we could use a well-crafted game genie code to make Ganon vulnerable to the boomerang (or other weapon) without affecting other parts of the game. To conform with the rules, the published AVI would involve Link going all the way up to Ganon, acting like the movie desyncs, and dying. However, those who view the run in an emulator would have the option of putting the code in before the final battle and see Link truly kill Ganon without a sword. This is actually the reason I contacted the Zelda Challenge author last week, and as I mentioned above he certainly wasn't much help. Another issue is that there's no guarantee that such a patch could be done only with an elegant single-byte change (a single game genie code). Requiring a whole handful of codes would take away the charm.
Actually, just now I thought up a cheesier solution. Try the same idea as above except use a code that turns the door above Ganon into a bomb-wall and have link just walk away from the final battle. Personally I think it would be a lot cooler to kill Ganon somehow, but at least this method is likely to entail only a single-line game genie code.
Experienced Forum User, Published Author, Player
(20)
Joined: 10/14/2005
Posts: 317
Hamelin wrote:
Hands down it has to be the most elaborate and best redo of Zelda 1 there is.
This hack definitely shows a lot of hard work, and by comparison it's not nearly as bad as his Zelda 2 hack... However, although this game features completely redone sprites and maps, it doesn't seem to have any asm mods and thus falls into the same low quality category as most hacks.
In fact, I contacted the author recently to get some help on Zelda patching, and it turns out he didn't even manage to edit damage attributes. I do hope his Zelda 3 Challenge hack might come out better, but even for skilled asm coders it's just really hard to make a good quality hack like Super Demo World: TLC.
